Docs & Writing

phoenixrr2113 Bundle

Deep research — clarify question, find primary sources, verify, deliver recommendation. Use when the answer requires checking documentation, papers, or official repos rather than guessing from memory.

gvkhosla Skill

Teaches you to read genuine PMF signals — and to stop fooling yourself with vanity metrics. Spawns 4 parallel agents analyzing retention, word-of-mouth, engagement depth, and revenue signals simultaneously. Use when build-cycle shows "Building" PMF signal, when you want an honest read on traction, or when you're not sure if what you're seeing is real. Produces pmf-assessment.md with a signal strength rating and the specific leading indicators to watch.

gvkhosla Skill

Stress-tests whether your problem is real, painful, and worth solving before you build anything. Runs 5 validation tests (max 25 points), then spawns 3 parallel research agents — Evidence Hunter, Market Sizer, and Alternative Mapper — to gather supporting data simultaneously. Use when you have a product idea and want to validate it's solving a genuine problem, or when you're unsure if people actually care. Produces problem-validation-report.md with a confidence score and clear go/no-go signal.

gvkhosla Skill

Reviews your product UI against Nielsen's 10 usability heuristics. Spawns 10 parallel agents — one per heuristic — to score, identify issues, and prescribe fixes simultaneously, then synthesizes into a severity-sorted report. Use after building your first screens, when users report confusion, or before showing the product to anyone important. Produces ux-review.md with prioritized issues and specific fixes.

gvkhosla Skill

Plans a 5-person user interview sprint and synthesizes findings. Use before building or whenever you are guessing about users instead of knowing. Produces research-report.md.

ryanthedev Bundle

Audits and improves code clarity and documentation using APOSD's obviousness rules: naming quality, comment accuracy, AI-facing doc completeness, and the comments-first workflow for new code. Not for generating project-wide convention docs (use code-standards) or structural simplification of the code itself (use aposd-simplifying-complexity).

agentscope-ai Skill

Answer user questions about QwenPaw installation and configuration: first locate and read local documentation, then distill the answer; if local information is insufficient, fall back to the official website documentation.

atman-33 Skill

Investigate a reported bug to its root cause — reproduce, trace the code path, identify the defect with evidence, assess blast radius, and propose fix candidates without changing code. Use when the user reports or pastes a bug/defect/unexpected behavior and wants it diagnosed, or asks "why is X happening".

qusong0627 Bundle

每日复盘后的股票推荐（多维度选股）— 综合 A股每日复盘（市场方向/板块/资金/新闻情绪）+ 个股深度分析（9层），用多维度筛选条件（L2微观结构为主/模型融合分/仓位信号/板块强度/新闻情绪）从全市场挑出未来几天大概率走强的股票，支持跨多日聚合（--window），输出候选榜 + 综合复盘 + Top 个股深度分析，落盘 PDF 到股票报告目录。用户说「选股」「推荐股票」「每日推荐」「明日看好」「选股推荐」「复盘后选股」时使用：跑复盘取数 → pick_candidates 多维打分 → Top N 深分 → 综合报告 → PDF。触发词：选股、股票推荐、每日推荐、明日看好、推荐股票、今日选股

qusong0627 Bundle

A股每日复盘（专业版）— 基于 QuantDB 本地数据 + 当日新闻情绪 + 模型推理信号 + L1/L2 因子截面 + 板块资金流的盘后复盘：指数、涨跌结构与涨停梯队、量能、行业/概念轮动、资金面、L2 微观结构、当日有新闻股票匹配、模型推理信号复盘（昨日推理命中率 + 明日信号Top5）、次日走势方向研判（六维信号加权 → 明确方向+置信度）。用户说「复盘」「每日复盘」「复盘某天」时使用：跑取数脚本（daily_review + news_review）→ 按模板写复盘报告 → 转 PDF → 落盘股票报告目录 → 聊天回复速览。触发词：复盘、每日复盘、今日复盘、盘后复盘、复盘20260814

qusong0627 Bundle

个股深度投研分析（智能体自主版）— 拉取 QuantMind 本地数据（371维特征/风险评分/模型推理分数/新闻）→ 多空子代理辩论 → 综合研判 → 生成 md 报告 → 导出 PDF 到平台「股票报告」页。任何大模型（deepseek/qwen/glm/openai/minimax）都能执行，不依赖容器投研管线。在 QuantBot / Claude Code 中深度分析股票时使用。触发词：投研分析、深度分析、个股分析、股票报告、生成报告、多空分析、AI分析师

qusong0627 Bundle

市场分析报告（大盘快照版）— 复用市场分析页面全部数据能力：核心指数、市场广度与情绪温度、行业板块/热门概念热力图、行业多日涨跌幅对比（1/3/5日）、板块资金流（1日/5日/10日）、个股主力资金 Top20、标签体系统计。用户说「市场分析」「大盘分析」「行情分析」「今天市场怎么样」「市场报告」时使用：跑取数脚本（market_analysis.py）→ AI 基于 facts 撰写解读 → Markdown → PDF（研报风）→ 落盘股票报告目录 → 聊天回复速览。触发词：市场分析、大盘分析、行情分析、市场报告、看盘、今天市场、今日市场、市场怎么样

qusong0627 Bundle

新闻情绪研究方法论 — Huntly RSS 42万篇历史新闻 → FinBERT+词典双引擎情绪 → 事件研究/来源预测力/时段特征/信号强度/首日动量/情绪反转/事件标签 七维深度分析 → 融合规律优化策略回测（来源白名单+时段过滤+多篇确认+首日动量+反转出场+连续/标签加成+无止损+动态止盈）→ 研报级 MD+PDF。用户说「新闻情绪」「新闻规律」「新闻策略」「情绪回测」「消息面研究」时使用。触发词：新闻情绪、新闻规律、情绪分析、消息面、新闻回测、情绪策略

fritzandfriends Bundle

**WORKFLOW SKILL** — Execute end-to-end ContosoUniversity migration benchmark: clear output folder, run L1 script + L2 Copilot transforms, build, run Playwright acceptance tests, and generate a numbered run report. WHEN: "run contoso migration test", "test contoso university migration", "contoso migration benchmark", "run CU migration". INVOKES: migration-toolkit (scripts + skills), dotnet CLI, Playwright tests.

fritzandfriends Bundle

**WORKFLOW SKILL** - Execute the end-to-end WingtipToys migration benchmark: clear samples\AfterWingtipToys, run the migration-toolkit against samples\WingtipToys, repair the generated app until Playwright acceptance tests pass, and write a numbered run report with embedded screenshots under dev-docs\migration-tests\wingtiptoys. WHEN: "run Wingtip migration", "test WingtipToys migration", "Wingtip benchmark", "migrate WingtipToys", "rerun Wingtip migration". INVOKES: migration-toolkit (scripts + skills), dotnet CLI, Playwright tests.

fritzandfriends Skill

Post-migration async/await and .NET 10 performance optimization pass for Blazor apps migrated from Web Forms. Applies modern runtime patterns after the app builds and runs. WHEN: "run L3 optimization", "apply async/await fixes", "optimize migrated Blazor app", "AsNoTracking queries", "StreamRendering", "IDbContextFactory pattern", "what .NET 10 optimizations can we apply", "generate L3 report".

mehrad-dm Bundle

report: a receipt for the work

notnotype Skill

汇报长任务交接、阻塞或当前交付证据。

openbkn-ai Bundle

生成阶段报告或最终归档报告。

orakitine Bundle

Transparent read-through cache for documentation lookups. Checks cache before WebFetch, caches results with expiration, garbage collects stale entries. TRIGGER when: fetching docs, API references, technical documentation, or user asks to look up library/framework usage. DO NOT TRIGGER when: user provides inline code or docs directly, or the information is already in context.

orakitine Bundle

Authors and maintains a living, executable engineering plan as a single Markdown document in docs/plans/. The plan has a fixed goal (Purpose + a Definition of Done) but a living path: a status-tracked phase/task checklist ([] idle, [wip] in progress, [x] done, [f] failed), an append-only amendments log, and metadata that survives across agents and sessions. Reads CONTEXT.md for domain vocabulary and links docs/adr/ decisions; never writes ADRs itself. Use when the user wants to plan, spec, or design new work; update or extend an existing plan; link plans/ADRs; or build (execute) a plan — or invokes /living-plan. Triggers: "plan this", "write a spec", "design doc", "living plan", "build the plan", "/living-plan".

orakitine Bundle

Reproduce and measure iOS-Safari-only layout, rendering, and interaction bugs on a booted iOS Simulator (iPad/iPhone) — the real-device counterpart to the desktop-Playwright browser-microscope. Drives Safari via safaridriver WebDriver + xcrun simctl to inject JS, read live geometry / computed styles / viewport units, and screenshot the device; uses macOS CGEvent drags to scroll and collapse the toolbar so dynamic-viewport dynamics appear. Use when a bug only shows in real iOS Safari and not in Playwright/Chromium/jsdom: dynamic viewport (dvh/svh/lvh) or stale initial-containing-block gaps, 100vh/100% sizing wrong after toolbar collapse, :focus-visible over-matching on programmatic or touch focus, momentum-scroll or rubber-band glitches, sticky/fixed drift, top-layer <dialog>/sheet coverage, safe-area insets, or any "works in Storybook but broken on iPad" report. Trigger phrases: "test on the iPad simulator", "reproduce this on iOS", "iOS Safari only bug", "check on real Safari".

r3bl-org Bundle

Run comprehensive Rust code quality checks including compilation, linting, documentation, and tests. Use after completing code changes and before creating commits.

r3bl-org Bundle

Writing Good Rust Documentation

srltas Bundle

Record an exploration note (PoC, review, code analysis, or any work without a JIRA issue) as a Markdown file in your public work-docs repo. Give it a free-form category and a slug and it scaffolds <category>/<YYYY-MM-DD>-<slug>.md from an exploration template (목적/배경/범위·방법/발견/결론/다음 단계/참고), fills it from the session, and commits it. While researching, actively use the collection's lookup/analysis skills (cubrid-manual, cmt-manual, and Understand-Anything if installed) so findings are verified, not asserted from memory. Complements the worklog skill, which is for issue-keyed work. The repo is PUBLIC, so keep internal-only detail in the master DOCX. Triggers on phrases like 'PoC 노트 남겨', '검토 노트 작성', '코드 분석 기록', 'record an exploration note', 'note this analysis'.

srltas Bundle

Record a piece of issue work as a Markdown note in your public work-docs repo, for a shareable, PR-linkable trail (the GitHub-visible record that complements the internal master DOCX). Give it a CUBRID Jira key (CBRD, TOOLS, APIS, CUBRIDQA, ...) or a Hibernate HHH key (or a topic slug) and it scaffolds <KEY>/<KEY>-<slug>.md from a consistent template (배경/원인/변경/검증/결과/링크), fills it from the session's work, and commits it. Pairs with jira-fetch for issue context. The repo is PUBLIC: keep internal-only detail in the master DOCX. Triggers on phrases like '작업 기록 남겨', 'worklog 작성', '이 이슈 문서로 정리해서 커밋', 'record this issue's work as markdown', 'publish a work note for CBRD-1234'.

srltas Bundle

Review SOMEONE ELSE'S pull request in depth, from a PR URL (or number) plus an optional Jira issue key. Fetches the whole repository locally as a detached git worktree (never touching your own branches or checkout), reads every changed file in full with no sampling, and establishes why the code is the way it is (git log, git blame, the PR body, the linked issue, prior PRs on the same files) before calling anything a defect: whatever it cannot ground is asked as a question to the author instead. Fans thirteen review passes out to parallel subagents, then merges them into one ranked Korean report with file:line evidence, split into 결함 / 판단 사항 / 범위 밖 / 질문. It reports to you and posts NOTHING: no review, no comment, no code edit. To judge review comments left on your own PR, use review-triage instead; to check your own unpushed work, use self-review. Triggers on phrases like '이 PR 리뷰해줘', 'PR 코드 리뷰 부탁해', '남의 PR 검토해줘', '이 풀리퀘 꼼꼼히 봐줘', 'review this PR', 'review someone else's pull request', 'deep review of PR 1629'.

srltas Skill

Draft a CUBRID JIRA issue (bug or task) with a concise English title and a Korean body under English section headers (Description as easy-to-read prose; the other sections in bullet style, 개조식). Use when you need to write up a CUBRID JIRA issue from the current work: a bug report (Description / Test Build / Repro / Expected / Actual / Additional Info) or a task/improvement (Description / Specification Changes / Implementation / Acceptance Criteria / Definition of Done). Produces a copy-paste draft only; it does not post to JIRA. It is the reverse of jira-fetch (which downloads an issue). Triggers on phrases like 'CUBRID JIRA 이슈 작성', 'jira 버그 리포트 초안', '이 작업 jira 이슈로 정리', 'draft a CUBRID jira issue', 'write a jira bug report'.

srltas Bundle

Run the Hibernate ORM test suite against CUBRID and measure a dialect change's impact: total failures, +N recovered / −N regressed vs a baseline, and failures grouped by signature/family. Use when verifying a CUBRIDDialect change, an auto_quote or capability-flag toggle, or any Hibernate-vs-CUBRID compatibility measurement. Pairs with the report skill (verify → numbers → 비교 분석 보고서). Triggers on phrases like 'dialect 검증', 'CUBRID 테스트 돌려', 'hibernate 스위트 실행', 'pass/fail 델타 측정', 'run the hibernate suite against CUBRID', 'measure dialect change'.

srltas Bundle

One-shot CUBRID JDBC driver verification + report. Builds the CUBRID JDBC driver, runs the CTP JDBC test suite against CUBRID, diffs pass/fail vs a baseline (a prior run's JSON, CTP result dir, or test-jdbc.xml), classifies broken cases (VerifyError, conversion, broker/env, …), and ALWAYS produces a Korean Word (.docx) 비교 분석 report. Use to verify a JDBC driver change or a Java-version/build-target change (e.g. v50 1.6 vs v52 1.8) end to end. Composes parse_ctp.py + the report skill. Triggers on phrases like 'JDBC 드라이버 검증', 'CTP 돌려', 'ctp jdbc 실행하고 리포트', 'verify the JDBC driver with CTP', 'build driver and run CTP'.

withwoz Skill

Show the WOZCODE savings report — calls saved, time saved, tokens saved, and lifetime totals.

withwoz Skill

Send feedback or a bug report to the WOZCODE team. Bug reports are tagged for triage and auto-attach session context (session id, anonymous id, OS, arch, Node version).

ar9av Bundle

Boots the generated game in headless Chromium with physics debug mode enabled, captures screenshots of all physics bodies, and emits a structured report of hitbox positions vs sprite positions. Use this to catch invisible hitbox misalignment bugs that only appear visually. Run in parallel with the other QA checks.

bagisto Bundle

Use when writing or updating any Bagisto documentation site — the developer documentation, the merchant user guide, or any other Bagisto docs repository — covering page content, code samples, screenshots, the sidebar, image naming, and moving or deleting pages. Trigger phrases include "docs", "documentation", "user guide", "developer documentation", "dev docs", "merchant documentation", "marketplace docs", "document this", "update the docs", "add a doc page", "screenshot", "ImagePopup", "redirect a doc page".

calebnewtonusc Bundle

Run the user's semester from a machine-readable ledger built out of their syllabi. Use when they mention a class, an assignment, an exam, a deadline, a professor, a reading, a grade, attendance, or a syllabus; when they ask what is due, what to work on, or whether they can miss a class; when they hand over a syllabus PDF; and before helping with any graded work, because each course sets its own rules about what help is allowed.

calebnewtonusc Bundle

Read from and write to the user's personal context repo, the private markdown store holding who they are, what they are working on, who they work with, and how their systems are wired. Use when the user asks what you know about something, states a durable fact worth persisting (new role, project shipped or died, collaborator joined, preference, infrastructure change), says "remember this", asks for a context review, or when substantial work just changed something that context should reflect. Also use before advising on anything personal, so the advice is grounded in their actual situation rather than a guess.